Long story short, all my SQL servers are doing nightly fulls even when they are specified to do an incremental. I have went through all the databases and ensured the Recovery Model is set to "Full" and I have also went through the SQL Server Agent Jobs / Management Jobs and ensured that there are any scheduled backup jobs there. However, all of my nsrsqlsv.log files read :
Promoting the backup level to full.[2224]
06/11/08 22:00:15 Last backup of database 'xxxxx' was not created using NMSQL Module.
Can anyone help me with this issue as I am chewing up tapes like crazy! The SQL Server logs also do not show any backup jobs in between the times I run the backups. I have also attempted to setup one of the servers to do interval backups every hour with "Force Incremental" enabled, but I still receive the above message in my nsrsqlsv.log
SQL incremental would work if SQL databases would be set in transaction logs mode else even if you set it incremental, it will take a FULL backup.
The only starneg thing is second part: Last backup was not created using NMSQL module. Is anyone running manual SQL backups? It could also be that it is not writing indexes properly and therefore it is not able to read its own databases for the last backup.
What is the status of backups in Networker and are you able to recover databases through Networker?
Yes, I can restore the databases and the backups run successfully. However, they run fulls every night. I do not see anything in the SQL History logs regarding other Backups. How do I ensure that the Databases are set in Transaction Logs mode?
Last time I've seen this was when SQL admins were also doing their dumps and did get recorded somewhere thus module knew last backup was not done by itself. To preserve log consistency full backup is performed then.
Within SQL you can schedule (or do it manually) dumps of database and if admin did it this may interfere with 3rd party backups. With newer systems, VSS could trigger backup too I believe.
Ok, I think I found the problem. It appears that the file level backup I do on the server at night is recording a log in the transaction log of SQL server which then screws up the SQL backup. How do I completely exclude all SQL file types from being included in the file level backup? I already have a MSSQL Exclude directive which skips .ldf and .mdf files, but are there more I must skip? Is there a guide to do this? I'm pretty positive this is what is causing the issue.
I have just confirmed that this is indeed the problem. How do I exclude SQL files from being backed up during my file level backup? Here is the directive that I am using: << "C:\" >> +skip: *.ldf *.LDF *.mdf *.MDF *.ndf *.NDF +aes: *.* +compressasm: .
I'm not sure what exactly is used by MS SQL so I really can't say exactly. I do know that newer releases of client for some reason would trigger MSDEWriter which might stand in the way (and may seem as SQL backup to application then). Not sure if this is addressed in latest releases (I know for a fact that with 7.2.2 client this is not the case).
It appears the my problem was in the way the directive was setup. I needed to have: << / >> +skip: *.ldf +skip: *.mdf +skip: *.ndf +skip: *.MDF +skip: *.LDF +skip: *.NDF +aes: *.* +compressasm: .
at the top of the directive rather than <<"C:\">> and so on.
amediratta
2 Intern
•
2014 Posts
889
0
Posted June 11th, 2008 21:00
The only starneg thing is second part: Last backup was not created using NMSQL module. Is anyone running manual SQL backups? It could also be that it is not writing indexes properly and therefore it is not able to read its own databases for the last backup.
What is the status of backups in Networker and are you able to recover databases through Networker?